PgCache vs Debezium and Redis

Debezium streams row changes out of Postgres. To keep Redis current with it, teams write a consumer that turns those changes into Redis updates. PgCache reads the same changes itself, with the cache built in, so there's no pipeline to build.

See which of your own queries PgCache would serve: check your workload or open a sample report.

Debezium and Redis facts checked against their docs on October 15, 2026.

PgCache

A caching proxy with change data capture built in. It reads the logical replication stream itself and keeps the data it has cached current.

One thing to run
No code that maps row changes to cache keys
Picks what to cache from live traffic

Debezium + Redis

Debezium captures row changes from Postgres and streams them to Kafka or Redis Streams. Code you write turns those changes into cache updates in Redis.

Where they differ

Topic	PgCache	Debezium + Redis
What you run	PgCache	Debezium, a stream such as Kafka, a consumer service and Redis
Code you write	None	A consumer that maps row changes to cache keys, plus cache code in the app
What it caches	SQL query results, picked from live traffic	Whatever your app and consumer put in Redis
Freshness	Within typical logical replication lag	Replication lag plus the pipeline’s own delay

Which one fits

Choose PgCache when

You want a cache kept current from Postgres changes, without building a pipeline
What you cache is the result of SQL queries

Choose Debezium and Redis when

You already run Debezium or Kafka
Changes also need to reach search, analytics or other services

What they have in common

Both keep a cache current from the Postgres replication stream
Neither speeds up writes

Common questions

Does PgCache use Debezium?

No. PgCache reads the Postgres logical replication stream itself.

Can I keep Debezium and add PgCache?

Yes. PgCache uses its own replication slot, so your other pipelines keep running.

What about Redis Data Integration?

Redis Data Integration is Redis’s packaged version of this pipeline. It runs on Debezium and is available with Redis Enterprise and Redis Cloud. Your app still reads from Redis instead of running SQL.